|Publication number||US7337356 B2|
|Application number||US 10/896,997|
|Publication date||Feb 26, 2008|
|Filing date||Jul 23, 2004|
|Priority date||Mar 20, 2003|
|Also published as||DE602004001869D1, DE602004001869T2, EP1604281A1, EP1604281B1, US7320091, US20050022094, US20050246613, WO2004084070A1|
|Publication number||10896997, 896997, US 7337356 B2, US 7337356B2, US-B2-7337356, US7337356 B2, US7337356B2|
|Inventors||Trevor Nigel Mudge, Todd Michael Austin, David Theodore Blaauw, Krisztian Flautner|
|Original Assignee||Arm Limited, University Of Michigan|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (66), Non-Patent Citations (2), Referenced by (32), Classifications (23), Legal Events (2)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a continuation of application Ser. No. 10/779,805, filed Feb. 18, 2004, now U.S. Pat. No. 7,162,661 which is a continuation-in-part of application Ser. No. 10/392,382, filed Mar. 20, 2003, the entire contents of which is hereby incorporated by reference in this application.
1. Field of the Invention
This invention relates to the field of integrated circuits. More particularly, this invention relates to the detection of errors including both random errors and systematic errors and the recovery from such errors within processing stages of an integrated circuit.
2. Description of the Prior Art
It is known to provide integrated circuits that can be considered to be formed of a series of serially connected processing stages (e.g. a pipelined circuit). Between each of the stages is a signal-capture element such as a latch or a sense amp into which one or more signal values are stored. The processing logic of each processing stage is responsive to input values received from preceding processing stages or elsewhere to generate output signal values to be stored in an associated output latch. The time taken for the processing logic to complete its processing operations determines the speed at which the integrated circuit may operate. If the processing logic of all stages is able to complete its processing operation in a short period of time, then the signal values may be rapidly advanced through the output latches resulting in high speed processing. The system cannot advance signals between stages more rapidly than the slowest processing logic is able to perform its processing operation of receiving input signals and generating appropriate output signals. This limits the maximum performance of the system.
In some situations it is desired to process data as rapidly as possible and accordingly the processing stages will be driven so as to advance their processing operations at as rapid a rate as possible until the slowest of the processing stages is unable to keep pace. In other situations, the power consumption of the integrated circuit is more important than the processing rate and the operating voltage of the integrated circuit will be reduced so as to reduce power consumption up to the point at which the slowest of the processing stages is again no longer able to keep pace. Both of these situations in which the slowest of the processing stages is unable to keep pace will give rise to the occurrence of processing errors (i.e. systematic errors).
One way of avoiding the occurrence of processing errors is to drive the integrated circuit with processing clocks having a frequency known to be less than the minimum permissible by a tolerance range that takes account of worst case manufacturing variation between different integrated circuits, operating environment conditions, data dependencies of the signals being processed and the like. In the context of voltage level, it is normal to operate an integrated circuit at a voltage level which is sufficiently above a minimum voltage level to ensure that all processing stages will be able to keep pace taking account of worst case manufacturing variation, environmental conditions, data dependencies and the like. It will be appreciated that the conventional approach is cautious in restricting the maximum operating frequency and the minimum operating voltage to take account of the worst case situations.
Besides systematic processing errors that result from the slowest of the processing stages being able to keep pace when a processor is run at too high a frequency or too low an operating voltage, integrated circuits are also subject to random errors known as single event upsets (SEUs). An SEU is a random error (bit-flip) induced by an ionising particle such as a cosmic ray or a proton in a device. The change of state is transient i.e. pulse-like so a reset or rewriting of the device causes normal behaviour thereafter. It is known to use error correction codes to detect and correct random errors. However, such error correction techniques necessarily introduce delay as a result of the processing time required for error detection and correction. This processing delay is justifiable in environments such as noisy communication channels where error rates are high yet it is important to suppress errors in the processed received data to within a predetermined error rate. By way of contrast, in the case of integrated circuits where it is generally desired to process data as rapidly as possible, it is undesirable to introduce error correction to critical paths of the data processing operations due to the delay and associated negative performance impact that error correction circuitry incurs.
Viewed from one aspect the present invention provides an integrated circuit for performing data processing, said integrated circuit comprising:
a plurality of processing stages, a processing stage output signal from at least one processing stage being supplied as a processing stage input signal to a subsequent processing stage, wherein said at least one processing stage comprises:
processing logic operable to perform a processing operation upon at least one coded input value to generate a processing logic output signal, said coded input value being an input value to which an error correction code has been applied;
a non-delayed signal-capture element operable to capture a non-delayed value of said processing logic output signal at a non-delayed capture time, said non-delayed value being supplied to said subsequent processing stage as said processing stage output signal following said non-delayed capture time;
a delayed signal-capture element operable to capture a delayed value of said processing logic output signal at a delayed capture time later than said non-delayed capture time;
error correction logic operable to detect an occurrence of a random error in said delayed value of said processing logic output signal, to determine if said detected random error is correctable using said error correction code and to either generate an error-checked delayed value or to indicate that said detected random error is not correctable;
a comparator operable to compare said non-delayed value with said error-checked delayed value to detect a change in said processing logic output signal at a time following said non-delayed capture time, said change being indicative of a systematic error whereby said processing logic has not finished said processing operation at said non-delayed capture time or of a random error in said non-delayed value; and
error-repair logic operable when said comparator detects said change in said processing logic output signal to perform an error-repair operation suppressing use of said non-delayed value either by replacing said non-delayed value by said error-checked delayed value in subsequent processing stages or by initiating repetition of said processing operation and processing operations of subsequent processing stages if said error correction logic indicates that said detected random error is not correctable.
The present technique recognises that the operation of the processing stages themselves can be directly monitored to find the limiting conditions in which they fail. When actual failures occur, then these failures can be corrected such that incorrect operation overall is not produced. The advantages achieved by the avoidance of excessively cautious performance margins in the previous approaches compared with the direct observation of the failure point in the present approach more than compensates for the additional time and power consumed in recovering the system when a failure does occur. Deliberately allowing such processing errors to occur such that critical paths fail to meet their timing requirements is highly counterintuitive in this technical field where it is normal to take considerable efforts to ensure that all critical paths always do meet their timing requirements.
Furthermore, the invention recognises that random errors in the delayed value may be detected and corrected by error correction logic deployed off the critical path of the data processing operations. Thus, when no systematic processing errors are detected by the comparator, the error correction logic has no adverse impact on the rapid progress of the computation. However, in the event that a processing error is in fact detected by the comparator, the delayed value available for use by the error repair logic to ensure forward progress of the computation is a reliable value on which a random error check and, where appropriate, random-error correction has been performed. Regardless of the presence of the error correction logic in the path of the delayed signal value, when processing errors are detected by the comparator, there will be a delay in the processing due to the need to perform the error-repair operation. Thus there is a surprising synergy between the provision of delayed signal-capture elements that enable repair of deliberately induced systematic processing errors and the application of error correction coding to correct random errors in the delayed signal values. The error correction logic provides the advantage of improving the reliability of the delayed value by detecting and correcting random errors without significantly delaying progress of the computation.
It will be appreciated that the processing operation performed by the processing logic could be a non-trivial processing operation that results in the value of the input signal changing relative to the value of the output signal, for example where the processing operation is a multiplication operation or a division operation with non-trivial operands. However, according to one preferred arrangement the processing operation performed by the processing logic is an operation for which the processing logic output signal is substantially equal to the processing stage input value when no errors occur in said processing operation.
For example, according to a first preferred arrangement the data processing operation that does not ordinarily change the input value could be read or write operation performed by a memory circuit. According to an alternative preferred arrangement the at least one processing stage is performed by a register and said processing operation is a read, write or move operation. According to a further alternative preferred arrangement in which the output signal value should be equal to the input signal value, the at least one processing stage is performed by a multiplexer and the processing operation is a multiplexing operation.
Whilst the present technique is applicable to both synchronous and asynchronous data processing circuits, the invention is well-suited to synchronous data processing circuits in which the plurality of processing stages are respective pipeline stages within a synchronous pipeline.
It will be appreciated that a variety of different error correction codes could be used to error correction encode the input value input value, for example, linear block codes, convolutional codes or turbo codes. However, for arrangements in which the output value is substantially equal to the input value of the processing logic, it is preferred that the input value is error correction encoded using a Hamming code and the error repair logic performs said correction and said detection using said Hamming code. Hamming codes are simple to implement and suitable for detecting and correcting single bit errors such as those typically resulting from SEUs.
Although some preferred arrangements involve value-preserving processing operations such as read/write operations and data moving operations, in alternative preferred arrangements the processing operation performed by the processing logic is a value-altering operation for which the processing logic output signal can be different from said processing stage input value even when no errors occur in said processing operation. Thus the present technique is suitable for application to processing logic elements such as adders, multipliers and shifters.
In arrangements where the processing operation is a value-altering processing operation, it is preferred that the input value is error correction encoded using an arithmetic code comprising one of: an AN code, a residue code, an inverse residue code or a residue number code. Such arithmetic codes facilitate detection and correction of random errors in processing operations involving arithmetic operators.
It will be appreciated that the comparator alone could be relied upon to detect the presence of systematic errors. However, in preferred arrangements the integrated circuit comprises a meta-stability detector operable to detect meta-stability in the non-delayed value and trigger the error-repair logic to suppress the use of the non-delayed value if found to be meta-stable.
Having detected the occurrence of a systematic error, via the comparator, there are a variety of different ways in which this may be corrected or compensated. In one preferred type of embodiment the error-recovering logic is operable to replace the non-delayed value with the error-checked delayed value as the processing stage output signal. The replacement of the known defective processing stage output signal with the correct value taken from the error-checked delayed value sample is strongly preferred as it serves to ensure forward progress through the data processing operations even though errors are occurring and require compensation.
A preferred arrangement is one in which the error-repair logic operates to force the delay value to be stored in the non-delay latch in place of the non-delayed value.
Whilst the present technique is applicable to both synchronous and asynchronous data processing circuits, the invention is well suited to synchronous data processing circuits in which the processing operations within the processing stages are driven by a non-delayed clock signal.
In the context of systems in which the processing stages are driven by the non-delayed clock signal, the error-repair logic can utilise this to facilitate recovery from an error by gating the non-delayed clock signal to provide sufficient time for the following processing stage to recover from input of the incorrect non-delayed value and instead use the correct error-checked delayed value.
In the context of embodiments using a non-delayed clock signal, the capture times can be derived from predetermined phase points in the non-delayed clock signal and a delayed clock signal derived from the non-delayed clock signal. The delay between the non-delayed capture and the delayed capture can be defined by the phase shift between these two clock signals.
The detection and recovery from systematic errors can be used in a variety of different situations, but is particularly well suited to situations in which it is wished to dynamically control operating parameters of an integrated circuit in dependence upon the detection of such errors. Counter intuitively, the present technique can be used to control operating parameters such that the system operates with a non-zero systematic error rate being maintained as the target rate since this may correspond to an improved overall performance, either in terms of speed or power consumption, even taking into account the measures necessary to recover from occurrence of both systematic and random errors.
The operating parameters which may be varied include the operating voltage, an operating frequency an integrated circuit body biased voltage (which controls threshold levels) and temperature amongst others.
In order to ensure that the data captured in the delayed latch is always correct, an upper limit on the maximum delay in the processing logic of any stage is such that at no operating point can the delay of the processing logic of any stage exceed the sum of the clock period plus the amount by which the delayed capture is delayed. As a lower limit on any processing delay there is a requirement that the processing logic of any stage should have a processing time exceeding the time by which the delayed capture follows the non-delayed capture so as to ensure that following data propagated along short paths does not inappropriately corrupt the delayed capture value. This can be ensured by padding short paths with one or more delay elements as required.
The present technique is applicable to a wide variety of different types of integrated circuit, such as general digital processing circuits, but is particularly well suited to systems in which the processing stages are part of a data processor or microprocessor.
In order to facilitate the use of control algorithms for controlling the operational parameters preferred embodiments include an error counter circuit operable to store a count of the detection of errors corresponding to a change in the delayed value compared with the non-delayed value. This error counter may be reached by software to carry out control of the operational parameters.
It will be appreciated that the delayed signal-capture element and non-delayed signal-capture element discussed above could have a wide variety of different forms. In particular, these may be considered to include embodiments in the form of flip-flops, D-type latches, sequential elements, memory cells, register elements, sense amps, combinations thereof and a wide variety of other storage devices which are able to store a signal value.
Viewed from another aspect, the present invention provides a method of controlling an integrated circuit for performing data processing, said method comprising the steps of:
supplying a processing stage output signal from at least one processing stage of a plurality of processing stages as a processing stage input signal to a subsequent processing stage, said at least one processing stage operating to:
perform a processing operation with processing logic upon at least one coded input value to generate a processing logic output signal, said coded input value being an input value to which an error correction code has been applied;
capturing a non-delayed value of said processing logic output signal at a non-delayed capture time, said non-delayed value being supplied to said subsequent processing stage as said processing stage output signal following said non-delayed capture time;
capturing a delayed value of said processing logic output signal at a delayed capture time later than said non-delayed capture time;
detect an occurrence of a random error in said delayed value of said processing logic output signal using error correction logic, to determine if said detected random error is correctable using said error correction code and to either generate an error-checked delayed value or to indicate that said detected random error is not correctable;
comparing said non-delayed value with said error-checked delayed value to detect a change in said processing logic output signal at a time following said non-delayed capture time, said change being indicative of a systematic error whereby said processing logic has not finished said processing operation at said non-delayed capture time or of a random error in said non-delayed value; and
when said change is detected, performing an error-repair operation using error-repair logic suppressing use of said non-delayed value either by replacing said non-delayed value by said error-checked delayed value in subsequent processing stages or by initiating repetition of said processing operation and processing operations of subsequent processing stages if said error correction logic indicates that said detected random error is not correctable.
The above, and other objects, features and advantages of this invention will be apparent from the following detailed description of illustrative embodiments which is to be read in connection with the accompanying drawings.
A meta-stability detector 7 serves to detect meta-stability in the output of the non-delayed latch 4, i.e. not at a clearly defined logic state. If such meta-stability is detected, then this is treated as an error and the value of the delay latch 6 is used instead.
On detection of an error, the whole pipeline may be stalled by gating the non-delayed clock signal 10 for an additional delayed period to give sufficient time for the processing logic in the following processing stage to properly respond to the corrected input signal value being supplied to it. Alternatively, it is possible that upstream processing stages may be stalled with subsequent processing stages being allowed to continue operation with a bubble inserted into the pipeline in accordance with standard pipeline processing techniques using a counterflow architecture (see the bubble and flush latches of
There are constraints relating to the relationship between the processing time taken by the processing logic within the processing stages and the delay between the non-delayed capture time and the delayed capture time. In particular, the minimum processing time of any processing stage should not be less than the delay in order to ensure that the delayed value captured is not corrupted by new data being outputted from a short delay processing stage. It may be necessary to pad short delay processing stages with extra delay elements to ensure that they do not fall below this minimum processing time. At the other extreme, it needs to be ensured that the maximum processing delay of the processing logic within a processing stage that can occur at any operational point for any operating parameters is not greater than the sum of the normal non-delayed operating clock period and the delay value such that the delay value captured in the delay value latch is ensured to be stable and correct.
There are a number of alternative ways in which the system may be controlled to tune power consumption and performance. According to one arrangement an error counter circuit (not illustrated) is provided to count the number of non-equal detections made by the comparator 6. This count of errors detected and recovered from can be used to control the operating parameters using either hardware implemented or software implemented algorithms. The counter is readable by the software. The best overall performance, whether in terms of maximum speed or lowest power consumption can be achieved by deliberately operating the integrated circuit with parameters that maintain a non-zero level of errors. The gain from operating non-cautious operating parameters in such circumstances exceeds the penalty incurred by the need to recover from errors.
According to an alternative arrangement, a hardware counter is provided as a performance monitoring module and is operable to keep track of useful work and of error recovery work. In particular, the counter keeps count of the number of useful instructions used to progress the processing operations being executed and also keeps count of the number of instructions and bubbles executed to perform error recovery. The software is operable to read the hardware counter and to use the count values to appropriately balance the overhead of error recovery and its effects on system performance against the reduced power consumption achieved by running the integrated circuit at a non-zero error rate.
At step 26 the processing logic from a stage i produces its output signal at a time Ti. At step 28 this is captured by the non-delayed latch and forms the non-delayed value. At step 30 the non-delayed value from the non-delayed latch starts to be passed to the following processing stage i+1 which commences processing based upon this value. This processing may turn out to be erroneous and will need recovering from should an error be detected.
Step 32 allows the processing logic to continue processing for a further time period, the delay time, to produce an output signal at time Ti+d. This output signal is latched in the delayed latch at step 34. The values within the delayed latch and the non-delayed latch are compared at step 36. If they are equal then no error has occurred and normal processing continues at step 37. If they are not equal, then this indicates that the processing logic at time Ti had not completed its processing operations when the non-delayed latch captured its value and started to supply that value to the subsequent processing stage i+1. Thus, an error condition has arisen and will require correction. At step 38 this correction is started by the forwarding of a pipeline bubble into the pipeline stages following stage i. At step 40 the preceding stages to stage i+1 are all stalled. This includes the stage i at which the error occurred. At step 42, stage i+1 re-executes its operation using the delayed latch value as its input. At step 44 the operating parameters of the integrated circuit may be modified as required. As an example, the operating frequency may be reduced, the operating voltage increased, the body biased voltage increased etc. Processing then continues to step 46.
If an insufficient number of errors is detected, then the operating parameter controlling circuits and algorithms can deliberately adjust the operating parameters so as to reduce power consumption and to provoke a non-zero error rate.
It will be seen that the sense amplifier 110 and the non-delayed latch 112 form part of the fast read mechanism. The sense amplifier 110 and the delayed latch 114 form part of the slow read mechanism. In most cases, the fast read result latched within the non-delayed latch 112 will be correct and no corrective action is necessary. In a small number of cases, the fast read result will differ from the slow read result latched within the delayed latch 114 and in this circumstance the slow read result is considered correct and serves to replace the fast read result with processing based upon that fast read result being suppressed. The penalty associated with a relatively infrequent need to correct erroneous fast read results is more than compensated for by the increased performance (in terms of speed, lower voltage operation, lower energy consumption and/or other performance parameters) that is achieved by running the memory 100 closer to its limiting conditions.
At step 126, the fast data read mechanism samples the value being output from the memory cell at that time. At step 128 this fast read data value is passed to subsequent processing circuits for further processing upon the assumption that it is correct. At step 130, the slow data reading mechanism samples a slow read data value. Step 132 compares the fast read value and the slow read value. If these are the same, then normal processing continues at step 134. However, if the sampled values are different, then step 136 serves to issue a suppression signal to the further circuits to which the fast read value has been passed and also to issue the slow read value in place of the fast read value to those further circuits such that corrective processing may take place.
Associated with each of the non-delayed latches 142 is a respective delayed latch 146. These delayed latches 146 serve to sample the signal value on the bus at a time later than when this was sampled and latched by the non-delayed latch 142 to which they correspond. Thus, a delay in the data value being passed along the bus for whatever reason (e.g. too low an operational voltage being used, the clock speed being too high, coupling effects from adjacent data values, etc) will result in the possibility of a difference occurring between the values stored within the non-delayed latch 142 and the delayed latch 146. The final stage on the pipeline bus 140 is illustrated as including a comparator 147 which compares the non-delayed value and the delayed value. If these are not equal, then the delayed value is used to replace the non-delayed value and the processing based upon the non-delayed value is suppressed such that the correction can take effect (the bus clock cycle may be stretched). It will be appreciated that these comparator and multiplexing circuit elements will be provided at each of the latch stages along the pipeline bus 140, but these have been omitted for the sake of clarity from
As the DSP circuit 144 does not itself support the non-delayed and delayed latching mechanism with its associated correction possibilities, it is important that the data value which is supplied to the DSP circuit 144 has been subject to any necessary correction. For this reason, an additional buffering latch stage 148 is provided at the end of the pipelined bus 140 such that any correction required to the data value being supplied to that latch and the attached DSP circuit 144 can be performed before that data value is acted upon by the DSP circuit 144. The buffering latch 148 can be placed in sufficient proximity to the DSP circuit 144 that there will be no issue of an insufficient available progation time etc. causing an error in the data value being passed from the buffering latch 148 to the DSP circuit 144.
It will be appreciated that the bus connections between the respective non-delayed latches 142 can be considered to be a form of processing logic that merely passes the data unaltered. In this way, the equivalence between the pipelined bus embodiment of
If the comparator 1024 detects a difference between the non-delayed signal value and the delayed signal value this indicates that either the processing operation was incomplete at the non-delayed capture time in the case that element 1014 represents processing logic or that the signal from the previous pipeline stage had not yet reached the present stage in the case of the element 1014 representing a data channel. In the event that such a difference is in fact detected, the value stored in the delayed latch 1018 is the more reliable data value since it was captured later when the processing operation is more likely to have been completed or the data from the previous stage is more likely to have arrived via the data channel. By supplying the result from the delayed latch to the next processing stage 1030 and suppressing use of the non-delayed value in subsequent processing stages, forward progress of the computation can be ensured. However, the reliability of the delayed signal value stored in the delayed latch 1018 can be compromised in the event that a single event upset occurred and corrupted the delayed value. The single event upset is effectively a pulse so it may well be missed by the non-delayed latch but picked up by the delayed latch. Such a single event upset will result in the comparator detecting a difference between the delayed and non-delayed values as a direct result of the single event upset and will then propagate the corrupted delayed value to subsequent processing stages. A single event upset that corrupts the non-delayed value will not be problematic since it will result in suppressing use of the erroneous non-delayed value and propagating the delayed value to subsequent stages.
The arrangement of
A given error correction code is capable of detecting a predetermined number of errors and of correcting a given number of errors. Thus the error detection module 1026 detects whether any errors have occurred and, if so, if the number of errors is sufficiently small such that they are all correctable. If correctable errors are detected then the signal value is supplied to the error correction module 1028 where the errors are corrected using the error correction code and the corrected delayed value is supplied to the comparator 1024. If it is determined by the comparator 1024 that the corrected delayed value differs from the non-delayed value then the error recovery procedure is invoked so that further propagation of the non-delayed value is suppressed in subsequent processing stages and the operations are instead performed using the corrected delayed value. On the other hand, if the comparator 1024 determines that the corrected delayed value is the same as the delayed value then there are two alternative possibilities for progressing the calculation. Firstly, the error recovery mechanism could nevertheless be invoked so that the non-delayed value is suppressed in subsequent processing stages and replaced by the corrected delayed value. Alternatively, since the non-delayed value is determined to have been correct (as evidenced by the equality of the non-delayed value and the corrected delayed value), the error recovery mechanism could be suppressed (despite the detection of an error in the delayed value) thus allowing the non-delayed value to continue to progress through the subsequent processing stages. However, if uncorrectable errors are detected in the delayed value by the error detection module 1026 then a control signal is supplied to suppress use of the corrupted delayed value. In this case forward progress of the computation cannot be achieved. The type of error correction encoding applied differs according to the nature of the channel/processing logic 1014.
Processing logic can be categorised as either value-passing or value-altering. Examples of processing logic that is value-passing are memory, registers and multiplexers. Examples of value-altering processing logic elements are adders, multipliers and shifters. Error detection and correction for value-altering processing logic elements is more complex than for value-passing processing logic elements because even when no error has occurred the value output by the logic stage 1014 is likely to be different from the input twelve-bit signal 1013.
As illustrated in
Arithmetic codes can be used to check arithmetic operators. Where
AN codes are arithmetic codes that involve multiplying the data word by a constant factor, for example a 3N code can be used to check the validity of an addition operation by performing the following comparison:
A further example of a class of arithmetic codes are residue codes, in which a residue (remainder of division by a constant) is added to the data bits as check bits e.g. a 3R code involves modulo (MOD) 3 operations and the following check is applied:
X MOD 3+Y MOD 3?=(X+Y)MOD 3
Consider the numerical example of X=14 and Y=7:
14 MOD 3=2 (codeword 111010, with last two bits as residue);
7 MOD 3=1 (codeword 011101);
and 21 MOD 3=0;
sum of residues MOD 3=(2+1) MOD 3=0=residue of (X+Y).
The present technique allows a more selective and indeed dynamic approach to be taken. A pipelined processing circuit 2000 includes delayed latches 2002 which can be used to detect the occurrence of errors in the signal values being captured by the non-delayed latches. The occurrence of these errors is fed back to a clock phase control circuit 204 which serves to adjust the relative phases of the clock signals being supplied to respective latches within the main path, i.e. the non-delayed latches. In this way, an adjustment is made whereby time is effectively borrowed from one processing stage and allocated to another processing stage. This may be achieved by tapping the clock signals to be used by the respective non-delayed latches from selectable positions within a delay line along which the basic clock signal is propagated.
The illustrated example, the processing logic between latch LA and latch LB is slower in operation than the processing logic in the subsequent stage. Accordingly, the clock signal being supplied to the non-delayed latch LB can be phase shifted so as to delay the rising edge of that clock signal (assuming rising edge latch capture) and thereby to extend the time available for the slow processing logic. This reduces the time available for the processing logic within the subsequent processing stage assuming that this is operating on the same basic clock signal as the other stage elements excluding the latch LB.
This timing balancing between processing stages can be performed dynamically during the ongoing operation of the circuit using feedback from the errors in operation detected using the delay latches. Alternatively, the balancing can be performed as a one-off operation during a manufacturing test stage or during a “golden boot” of the integrated circuit. The delayed latches shown in
It is important that errant pipeline results not be written to architectured state before it has been validated by the comparator. Since validation of delayed values takes two additional cycles (i.e., one for error detection and one for panic detection), there must be two non-speculative stages between the last delayed latch and the writeback (WB) stage. In our design memory accesses to the data cache are non-speculative, hence, only one additional stage labelled ST for stabilise is required before writeback (WB). The ST stage introduces an additional level of register bypass. Since store instructions must execute non-speculatively, they are performed in the WB stage of the pipeline.
In aggressively clocked designs, it may not be possible to implement global clock gating without significantly impacting processor cycle time. Consequently, a fully pipelined error recover mechanism based on counterflow, pipelining techniques has been implemented. The approach illustrated in
Recall from the description of
A key requirement of the pipeline recover control is that it not fail under even the worst operating conditions (e.g. low voltage, high temperature and high process variation). This requirement is met through a conservative design approach that validates the timing of the error recovery circuits at the worst-case subcritical voltage.
Although illustrative embodiments of the invention have been described in detail herein with reference to the accompanying drawings, it is to be understood that the invention is not limited to those precise embodiments, and that various changes and modifications can be effected therein by one skilled in the art without departing from the scope and spirit of the invention as defined by the appended claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US3893070 *||Jan 7, 1974||Jul 1, 1975||Ibm||Error correction and detection circuit with modular coding unit|
|US3905023||Aug 15, 1973||Sep 9, 1975||Burroughs Corp||Large scale multi-level information processing system employing improved failsaft techniques|
|US4339657 *||Feb 6, 1980||Jul 13, 1982||International Business Machines Corporation||Error logging for automatic apparatus|
|US4633465||Mar 27, 1985||Dec 30, 1986||At&T Bell Laboratories||Eye violation detectors|
|US4669092 *||Sep 26, 1985||May 26, 1987||U.S. Philips Corporation||Arrangement for receiving digital data comprising an arrangement for adaptive timing recovery|
|US4756005 *||May 27, 1986||Jul 5, 1988||American Telephone And Telegraph Company At&T Bell Laboratories||Digital signal regenerator arranged for fault location|
|US4918709||Dec 5, 1988||Apr 17, 1990||U.S. Philips Corporation||Data demodulator baud clock phase locking|
|US4975930||Nov 2, 1988||Dec 4, 1990||Digital Equipment Corporation||Digital phase locked loop|
|US5043990 *||Dec 2, 1988||Aug 27, 1991||Hitachi, Ltd.||Semiconductor integrated circuit device|
|US5203003||Mar 28, 1991||Apr 13, 1993||Echelon Corporation||Computer architecture for conserving power by using shared resources and method for suspending processor execution in pipeline|
|US5276690 *||Jan 30, 1992||Jan 4, 1994||Intel Corporation||Apparatus utilizing dual compare logic for self checking of functional redundancy check (FRC) logic|
|US5291496 *||Oct 18, 1990||Mar 1, 1994||The United States Of America As Represented By The United States Department Of Energy||Fault-tolerant corrector/detector chip for high-speed data processing|
|US5313625||Jul 30, 1991||May 17, 1994||Honeywell Inc.||Fault recoverable computer system|
|US5321705 *||Mar 30, 1992||Jun 14, 1994||Motorola, Inc.||Error detection system|
|US5400370||Feb 24, 1993||Mar 21, 1995||Advanced Micro Devices Inc.||All digital high speed algorithmic data recovery method and apparatus using locally generated compensated broad band time rulers and data edge position averaging|
|US5402273||Oct 1, 1992||Mar 28, 1995||Conner Peripherals, Inc.||Circuit for determining window margin for data transitions in a streaming data device|
|US5408200||Feb 8, 1994||Apr 18, 1995||Storage Technology Corporation||Intelligent phase detector|
|US5426746||Jul 8, 1994||Jun 20, 1995||Kabushiki Kaisha Toshiba||Microcontroller with program recomposing function|
|US5455536 *||Jan 12, 1994||Oct 3, 1995||Nec Corporation||Demodulator circuit and demodulating method employing bit error rate monitor|
|US5463351||Sep 29, 1994||Oct 31, 1995||Motorola, Inc.||Nested digital phase lock loop|
|US5504859 *||Nov 9, 1993||Apr 2, 1996||International Business Machines Corporation||Data processor with enhanced error recovery|
|US5528637 *||Oct 12, 1994||Jun 18, 1996||Alcatel N.V.||Synchronizing circuit|
|US5553232||Jun 13, 1994||Sep 3, 1996||Bull Hn Informations Systems Inc.||Automated safestore stack generation and move in a fault tolerant central processor|
|US5572662 *||May 31, 1995||Nov 5, 1996||Fujitsu Limited||Data processing apparatus|
|US5625652||Jun 7, 1995||Apr 29, 1997||Pacific Communication Sciences, Inc.||Timing recovery controller and method for adjusting the timing of synchronizing windows in a PSK demodulator|
|US5627412||Jun 7, 1995||May 6, 1997||Norand Corporation||Dynamically switchable power supply|
|US5737369 *||Feb 28, 1996||Apr 7, 1998||Motorola, Inc.||Apparatus and method for recovering data in the presence of error transients|
|US5859551 *||Apr 29, 1997||Jan 12, 1999||Fujitsu Limited||Digital PLL circuit|
|US5870446 *||Mar 11, 1996||Feb 9, 1999||Adtran, Inc.||Mechanism for automatically adjusting the phase of a transmission strobe clock signal to correct for misalignment of transmission clock and data signals|
|US5914903||Jan 30, 1998||Jun 22, 1999||Fujitsu Limited||Semiconductor memory device|
|US6067256||May 5, 1998||May 23, 2000||Mitsubishi Denki Kabushiki Kaisha||Static semiconductor memory device operating at high speed under lower power supply voltage|
|US6114880 *||Mar 14, 1997||Sep 5, 2000||Philips Semiconductor Vlsi, Inc.||Dynamic over frequency detection and protection circuitry|
|US6167526||Dec 29, 1998||Dec 26, 2000||Adaptec, Inc.||Method and apparatus for synchronizing a decoder circuit with a phase-encoded data signal in a data storage device|
|US6173423 *||Feb 9, 1998||Jan 9, 2001||Bull, S.A.||Device and process for detecting errors in an integrated circuit comprising a parallel-serial port|
|US6188610||Feb 3, 2000||Feb 13, 2001||Kabushiki Kaisha Toshiba||Electrically erasable and programmable nonvolatile semiconductor memory device having data holding function and data holding method|
|US6222660 *||Jun 9, 1998||Apr 24, 2001||Tektronix, Inc.||Adaptive power supply for avalanche photodiode|
|US6476643||Oct 5, 2001||Nov 5, 2002||Stmicroelectronics S.A.||Asynchronous circuit for detecting and correcting soft error and implementation method thereof|
|US6523201 *||May 4, 1999||Feb 25, 2003||Eliza A. De Michele||Sleep support system|
|US6693985 *||Oct 26, 2001||Feb 17, 2004||Silicon Image||Clock and data recovery method and apparatus|
|US6741110 *||May 28, 2002||May 25, 2004||Lsi Logic Corporation||Method and/or circuit for generating precision programmable multiple phase angle clocks|
|US6772388 *||Dec 6, 2000||Aug 3, 2004||Motorola, Inc||Apparatus and method for providing optimal adaptive forward error correction in data communications|
|US6799292 *||Mar 7, 2001||Sep 28, 2004||Matsushita Electric Industrial Co., Ltd.||Method for generating test pattern for semiconductor integrated circuit and method for testing semiconductor integrated circuit|
|US6834367 *||Dec 21, 2000||Dec 21, 2004||International Business Machines Corporation||Built-in self test system and method for high speed clock and data recovery circuit|
|US6907553 *||Jun 18, 2001||Jun 14, 2005||Lsi Logic Corporation||Method and apparatus for estimation of error in data recovery schemes|
|US6944468 *||Jun 28, 2002||Sep 13, 2005||Ntt Docomo, Inc.||Transmit power control method and transmit power control system suitable to mobile communications|
|US6958627||Sep 21, 2001||Oct 25, 2005||Trustees Of Columbia University In The City Of New York||Asynchronous pipeline with latch controllers|
|US6977910 *||Dec 31, 1998||Dec 20, 2005||Texas Instruments Incorporated||Power control with space time transmit diversity|
|US7010074 *||Jan 11, 2002||Mar 7, 2006||Nec Electronics Corporation||Oversampling clock recovery having a high follow-up character using a few clock signals|
|US7046056 *||Apr 26, 2005||May 16, 2006||Rambus Inc.||System with dual rail regulated locked loop|
|US7085993 *||Jul 29, 2002||Aug 1, 2006||International Business Machine Corporation||System and method for correcting timing signals in integrated circuits|
|US7096137 *||Dec 2, 2003||Aug 22, 2006||Silverbrook Research Pty Ltd||Clock trim mechanism for onboard system clock|
|US7116744 *||Oct 18, 2001||Oct 3, 2006||Fujitsu Limited||Clock recovery circuit and receiver circuit for improving the error rate of signal reproduction|
|US7142623 *||May 31, 2002||Nov 28, 2006||International Business Machines Corporation||On-chip system and method for measuring jitter tolerance of a clock and data recovery circuit|
|US7162661 *||Feb 18, 2004||Jan 9, 2007||Arm Limited||Systematic and random error detection and recovery within processing stages of an integrated circuit|
|US20010016927||Apr 26, 2001||Aug 23, 2001||Poisner David I.||Method and apparatus for verifying that data stored in a memory has not been corrupted|
|US20020038418||Sep 25, 2001||Mar 28, 2002||Matsushita Electric Industrial Co., Ltd.||Computer system|
|EP0366331A2||Oct 17, 1989||May 2, 1990||Advanced Micro Devices, Inc.||System and method for error detection in the result of an arithmetic operation|
|EP0374420A2||Oct 21, 1989||Jun 27, 1990||International Business Machines Corporation||A residue checking apparatus for detecting errors in add, substract, multiply, divide and square root operations|
|EP0653708A2||Oct 12, 1994||May 17, 1995||Hitachi, Ltd.||Logic circuit having error detection function, redundant resource management method, and fault tolerant system using it|
|JP2001175542A||Title not available|
|JPS6020398A||Title not available|
|JPS6224498A||Title not available|
|SU809350A1||Title not available|
|WO2000054410A1||Mar 8, 2000||Sep 14, 2000||Centre Nat Rech Scient||Logic circuit protected against transitory perturbations|
|WO2001046800A2||Dec 19, 2000||Jun 28, 2001||Gen Instrument Corp||Dual-mode processor|
|WO2004084072A1||Feb 26, 2004||Sep 30, 2004||Advanced Risc Mach Ltd||Error detection and recovery within processing stages of an integrated circuit|
|1||"ARM710 Data Sheet" Dec. 1994, Advanced RISC Machines Ltd. (ARM).|
|2||N. Kanekawa et al, "Fault Detection and Recovery Coverage Improvement by Clock Synchronized Suplicated Systems with Optimal Time Diversity" Fault-Tolerant Computing, Jun. 1998, pp. 196-200.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7650551 *||Aug 16, 2007||Jan 19, 2010||Arm Limited||Error detection and recovery within processing stages of an integrated circuit|
|US7701240 *||Dec 13, 2005||Apr 20, 2010||Arm Limited||Integrated circuit with error correction mechanisms to offset narrow tolerancing|
|US7895469 *||Oct 14, 2008||Feb 22, 2011||Arm Limited||Integrated circuit using speculative execution|
|US8060814||Aug 21, 2009||Nov 15, 2011||Arm Limited||Error recovery within processing stages of an integrated circuit|
|US8161367||Oct 7, 2008||Apr 17, 2012||Arm Limited||Correction of single event upset error within sequential storage circuitry of an integrated circuit|
|US8171386||Mar 27, 2008||May 1, 2012||Arm Limited||Single event upset error detection within sequential storage circuitry of an integrated circuit|
|US8185786 *||Oct 13, 2010||May 22, 2012||Arm Limited||Error recovery within processing stages of an integrated circuit|
|US8185812||Dec 11, 2006||May 22, 2012||Arm Limited||Single event upset error detection within an integrated circuit|
|US8369883 *||Jan 19, 2007||Feb 5, 2013||Nec Corporation||Transmission power control technique and wireless communications system using the same|
|US8381009 *||Aug 3, 2006||Feb 19, 2013||Freescale Semiconductor, Inc.||Device and method for power management|
|US8407537||Oct 13, 2010||Mar 26, 2013||Arm Limited||Error recover within processing stages of an integrated circuit|
|US8493120||Mar 10, 2011||Jul 23, 2013||Arm Limited||Storage circuitry and method with increased resilience to single event upsets|
|US8621272 *||Dec 29, 2008||Dec 31, 2013||Arm Limited||Integrated circuit with error repair and fault tolerance|
|US8650470||Oct 25, 2010||Feb 11, 2014||Arm Limited||Error recovery within integrated circuit|
|US8862935||Dec 30, 2013||Oct 14, 2014||Arm Limited||Integrated circuit with error repair and fault tolerance|
|US9021298||Dec 30, 2013||Apr 28, 2015||Arm Limited||Integrated circuit with error repair and fault tolerance|
|US9164842||Jun 25, 2013||Oct 20, 2015||Arm Limited||Error recovery within integrated circuit|
|US9244123||Nov 25, 2014||Jan 26, 2016||Freescale Semiconductor, Inc.||Synchronous circuit, method of designing a synchronous circuit, and method of validating a synchronous circuit|
|US20060200699 *||Dec 13, 2005||Sep 7, 2006||Arm Limited||Integrated circuit with error correction mechanisms to offset narrow tolerancing|
|US20070162798 *||Dec 11, 2006||Jul 12, 2007||Arm Limited||Single event upset error detection within an integrated circuit|
|US20070173279 *||Jan 19, 2007||Jul 26, 2007||Nec Corporation||Transmission power control technique and wireless communications system using the same|
|US20070288798 *||Aug 16, 2007||Dec 13, 2007||Arm Limited||Error detection and recovery within processing stages of an integrated circuit|
|US20090106616 *||Oct 14, 2008||Apr 23, 2009||Arm Limited||Integrated circuit using speculative execution|
|US20090249175 *||Mar 27, 2008||Oct 1, 2009||Arm Limited||Single Event Upset error detection within sequential storage circuitry of an integrated circuit|
|US20100019818 *||Aug 3, 2006||Jan 28, 2010||Freescale Semiconductor Inc.||Device and method for power management|
|US20100058107 *||Aug 21, 2009||Mar 4, 2010||Arm Limited||Error recovery within processing stages of an integrated circuit|
|US20100088565 *||Oct 7, 2008||Apr 8, 2010||Arm Limited||Correction of single event upset error within sequential storage circuitry of an integrated circuit|
|US20100275080 *||Dec 29, 2008||Oct 28, 2010||Shidhartha Das||Integrated circuit with error repair and fault tolerance|
|US20110093737 *||Apr 21, 2011||Krisztian Flautner||Error recovery within processing stages of an integrated circuit|
|US20110107166 *||May 5, 2011||Arm Limited||Error recovery within integrated circuit|
|US20110126051 *||May 26, 2011||Krisztian Flautner||Error recover within processing stages of an integrated circuit|
|WO2012164541A1||Jun 1, 2012||Dec 6, 2012||Politecnico Di Torino||Method and circuit for solving metastability conditions and recovering signal errors in digital integrated circuits|
|U.S. Classification||714/10, 714/30, 714/E11.056, 714/E11.064, 714/819, 714/55, 714/708|
|International Classification||G06F11/00, G06F11/16, G06F11/10|
|Cooperative Classification||G06F11/0721, G06F11/0793, G06F9/3861, G06F9/3869, G06F11/104, G06F11/1608, G06F11/167, G06F11/1695, G06F11/183|
|European Classification||G06F11/16Z, G06F11/16M2, G06F9/38P2, G06F9/38H|
|Jul 27, 2011||FPAY||Fee payment|
Year of fee payment: 4
|Aug 12, 2015||FPAY||Fee payment|
Year of fee payment: 8